This September, Tune Out ‘Worst Month’ Noise

With summer unofficially about to turn to autumn, many minds are on football season’s start, changing leaves and, we guess, pumpkin-spiced everything. But many pundits are fixated on something else: September, so the old adage goes, is the worst month for stocks. This year, fallout fear is especially pronounced, given sour sentiment and stocks’ steep post-correction rally. But before you fall prey to this “worst-month” chatter, remember: September is just a month, not a market driver. You are best blocking out seasonal market noise.

Mind you, the old adage is technically true. Since 1926, September’s average return of -0.8% is indeed the worst of any calendar month.^[i] And stocks have fallen four of the last five Septembers, which could be contributing to today’s buzz.^[ii] Coming amid widespread worries over tariffs, high valuations and allegedly narrow market breadth, many see the approach of the calendar’s worst month as a troubling sign.

But there are some considerable issues with this “analysis.” To start, September’s negative average return is tied to a few lofty outliers. This includes 1931’s -29.6% fall (the S&P 500’s worst monthly return), 2008’s -8.9% and 2022’s -9.4% drop—among others.^[iii]

Simple average returns can relate a skewed story. One way to remove skew on both sides is by using median returns, or the midpoint of all monthly returns. And for September, this figure is essentially flat.^[iv] Not great, but not negative, either. Oh, and from a frequency standpoint, September returns are positive 52% of the time—slightly more often than not.^[v]

There are other, more substantial issues, though. Namely, correlation doesn’t mean causation—as these large monthly drops illustrate. They didn’t happen because it was September. For example, take September 1931 (please). Amid the Great Depression, serial bank runs put immense pressure on the American banking system, hammering US markets. This pain wasn’t reflected solely in September, nor did it begin then. This period included many poor months for the S&P 500, including May 1931’s -13.2%, December 1931’s -13.8% and April 1932’s -19.8%.^[vi] The calendar triggered none of these. The Great Depression did.

Or consider chaotic September 2008, when the Federal Reserve forced Lehman Brothers to fail while the US government haphazardly responded to the global financial crisis. Here, too, poor S&P 500 monthly returns didn’t start with September—the bear market began almost one year earlier, in October 2007. The downturn featured several negative months, including October 2008, when stocks fell -16.8%—nearly twice as much as the previous September. February 2009’s -10.6% was another double-digit drop that followed.^[vii]

More recently, September 2022 fell near the end of a shallow bear market driven by a cornucopia of fears including recession, hot inflation, Fed rate hikes, Russia’s war in Ukraine and more. Herein lies the point: None of these drivers have anything to do with September as a month. Or Labor Day. Or, dare we say, pumpkin spice. The economic, political and sentiment drivers at the time explain them much better. Heck, 8 of the worst 10 Septembers came amid pre-existing bear markets.^[viii] Those cyclical moves explain the bad months, not the other way around.

More broadly speaking, seasonality just isn’t predictive. This becomes apparent when you look at other adages, like “Sell in May,” which urges you to dodge the “weak” summer months, returning on Halloween or St. Leger Day (a mid-September British horse race). But … this year selling on April 30 would have meant missing 14.8% S&P 500 gains thus far.^[ix] Last year, US markets rose 12.3% over this period.^[x] This doesn’t appear worth trying to dodge.

Importantly, these aren’t outliers. Historically, summers have been fine for stocks, averaging positive returns. Thus, by following the Sell in May myth, you risk losing out on solid gains. Over time, these mistakes can pile up and cause you to veer from your long-term goals and objectives. This is the chief risk falling prey to seasonality presents.

All in all, calendars just aren’t reliable market indicators. While stocks fell in four of the past five Septembers, they rose 2.0% last year.^[xi] And they didn’t drop in any of the four Septembers from 2016 to 2019.^[xii] Point being, all seasonal adages work some of the time, helping confirm the biases around them and lending to headlines’ chatter.

But there is no market-predictive quality to the calendar—stocks aren’t crops, rotating with the season. The market is a complex and adaptive system that pre-prices widely known information. Time flies, but we doubt many are unaware the calendar is about to flip to September.